World Targets $40B AI Fraud Risk With Bank-Grade Human ID Suite

World, the identity project co-founded by Sam Altman, has launched World ID for banking and finance, a suite of tools that lets financial institutions confirm a real person sits behind transactions, trades and customer requests, the company announced on Sunday.

Bain Capital, Pantera Capital, VanEck Funds, OpenExchange and Outtake are the initial design partners, according to World. The rollout targets a fraud problem with hard numbers attached: Deloitte forecasts that AI-enabled fraud losses in the U.S. could reach $40 billion by 2027, up from $12.3 billion in 2023.

How does the verification stack work?

The offering rests on three cryptographic proofs delivered through the World ID app:

• Proof of Human — establishes that a user is a real, unique person via a one-time visit to one of World's iris-scanning Orb devices.
• Proof of Presence — runs a selfie liveness check on a phone.
• Proof of Attributes — confirms details such as nationality or age from a government ID, revealing only the requested information.

Banks could require a verified human to approve ACH and wire transfers, or to recover an account, without collecting documents or storing additional personal data, World said. A developer toolkit called AgentKit, still in beta, lets users delegate certain proofs to AI agents so the agents can transact on a person's behalf.

The suite also includes World ID for Zoom, extending earlier integrations with Zoom and Tinder. Participants verified at an Orb device get a "Human Badge" on their video tile after a live selfie check, and hosts will soon be able to require verification in the waiting room before a call starts.

Why now?

World flagged the case of an employee at engineering firm Arup who wired roughly $25.6 million in early 2024 after joining a video call in which every other participant was a deepfake. Since that incident, AI-driven impersonation has grown more convincing, the company said, straining fraud controls at financial institutions.

Adoption numbers give the pitch some scale. More than 18 million people hold a Proof of Human ID, according to World, and the World Foundation has raised $52.5 million this year to expand the network.

What about the regulatory record?

The same iris-scanning model that anchors Proof of Human has drawn enforcement action across multiple jurisdictions over biometric data handling:

• Hong Kong's privacy watchdog ordered World to halt iris-scanning operations in 2024.
• South Korea's privacy regulator fined the project $830,000 later that year.
• Spain's data protection agency ordered the deletion of all iris scan data in December 2024.
• A Kenyan court ordered the project to delete biometric user data last year.
• Brazil's data protection authority denied World's appeal against a national ban.

That record matters for the banking push. Any financial institution adopting Proof of Human would depend on a biometric collection pipeline that regulators in at least five markets have restricted or penalized, a compliance question the design partners will have to weigh against the fraud losses the product claims to prevent.

The launch stakes out World's position in a market where AI agents and deepfakes are eroding default assumptions about who — or what — is on the other side of a transaction, and where banks now face a 2027 deadline of sorts in Deloitte's $40 billion loss projection.